VLAN - (b) BVI - Semantic Success: leaderboard
Full1931trajectorysplit; user-conditionedlanguageinstruction. SRlan requiresalluser-specificsemanticconstraints; IAS weightedstructuredsimilaritytocanonicaltarget. Seven releasedVLMs; exactInstructlabelsretained; noactionSRderivedmetricused. Metric: Native score (0–1). Source: https://arxiv.org/html/2611.18346v1
Metric: Native score (0–1). Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 8B Instruct | 0.436 | #1025 |
| 2 | InternVL3.5-8B | 0.375 | #1588 |
| 3 | Qwen 2.5 VL 7B Instruct | 0.276 | #1893 |
| 4 | Qwen 2.5 VL 3B Instruct | 0.265 | #2648 |
| 5 | Llama 3.2 11B Instruct | 0.183 | #5226 |
| 6 | Molmo2-8B | 0.131 | #1629 |
| 7 | Idefics3-8B-Llama3 | 0.118 | #5775 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=vlan-b-bvi-semantic-success · How It Works · Data refreshed daily, snapshot 2026-11-17.